CONTROL OF HEAD-FORM IN PLAXARIA 119 



conditions in other parts of the individual. In short, the forma- 

 tion of a new head on an isolated piece of Planaria, or for that 

 matter, of any other form is a process determined primarily 

 by the constitution of the embryonic cells concerned and not by 

 correlative conditions in other parts. The fact that a head 

 arises directly at the cut surface, whatever the level of the body 

 from which the piece is taken is of itself very strong evidence 

 that this is the case, for if correlative factors in the piece deter- 

 mined the differentiation of the new tissue we should expect 

 that the first parts to appear at the anterior end would be those 

 which in nature are next anterior to this level and that as these 

 are successively formed the head would finally appear. As a 

 matter of fact, however, the head appears first and the other 

 parts arise by redifferentiation of regions of the piece into more 

 anterior regions under the influence of the new head. In short, 

 the piece does not determine the formation of a new head at its 

 anterior end although it may retard or inhibit the process. 

 The developing head, however, does determine the reorganiza- 

 tion and redifferentiation of other parts of the piece into those 

 regions of the body which normally lie between the head and 

 the level of the piece. The influence of y upon x is merely nega- 

 tive or inhibitory, while the influence of x upon y is positive and 

 determining. I have pointed out elsewhere (Child, 'lie, '14 b, 

 '15 b) how these facts fall into Hne \\i\h the general conceptions 

 of the axial gradient and of physiological dominance and sub- 

 ordination. Stating the case in terms of these conceptions, 

 we may say that if rate y is sufficiently high y dominates x, 

 inhibits its independent development as a head, and if its dom- 

 inance is sufficienth' complete, which is rarely the case in Planaria 

 dorotocephala. may even determine the development into 

 a posterior end. If, however, rate x is sufficiently high in re- 

 lation to rate y, x becomes the dominant region, develops in- 

 dependently of y and determines its redifferentiation. 



Turning now to the cyanide experiments we see that they are 

 not only readily interpreted on this basis but afford very strong 

 evidence for this conception of the process of reconstitution. 

 KCN as a reagent which depresses metabolism, inhibits to a 



